"Dave Plowman (News)" <
da...@davenoise.co.uk> wrote in message
news:533546b...@davenoise.co.uk...
> In article <op.wut0s...@duncan-tosh.lan>,
> Duncan Wood <
nnt...@dmx512.co.uk> wrote:
>> Google alternator load dump,
> I'd rather not, since so much is speculation.
I took the trouble to have a look at the top reference, which was the Wiki
one. It seems to contradict itself line by line, but the basic premise is
that if the battery is *disconnected* whilst the alternator is trying to
charge it, an over-voltage condition will be produced due to the fact that
the alternator control circuitry had the field windings being supplied with
the appropriate current to keep the alternator 'wide open' in order to be
supplying the maximum charging current to the battery. The contention is
that the alternator will continue to generate a large current that now has
nowhere to go, other than into other things attached to the battery, for up
to 400 mS before the field winding current is reduced by the ECU to
'throttle back' the alternator. One such connected item is the ECU itself,
and it is implied that this large current which continues to be generated,
can cause a peak voltage of up to 120 volts to be applied to the ECU, which
is what kills it.
Well, quite apart from the pseudo-science in all that, where the author
seems to be mixing current and voltage willy nilly to arrive at a
conclusion, two things come to mind. First, unless the receiver battery has
an open circuit cell, disconnecting the feed from the donor vehicle should
have little effect, as by far, the largest load on the alternator will be
the flat receiver battery, which will continue to drag as much current from
its own alternator as it will let it have. The donor battery / alternator
should not have any effect on the current demand of the receiver battery
from its own alternator, as the donor is a net exporter of current to the
flat receiver battery.
Secondly, the article does go on to say that there are published spike shape
specifications for the electronics designers to design against, and it
mentions varistors and similar protection devices that could be employed.
These are not esoteric devices. They are in extremely common usage right
across the field of electronics design from domestic to industrial, and are
employed anywhere that a troublesome voltage spike is likely to occur. They
are fast and extremely good at what they do, and above all, are cheap. Using
one, or a similar device, to protect the ECU against a 120v pulse on its
supply rail would, IMHO, be the very least that a good designer would do to
protect his 'delicate' circuitry from the extremely electrically hostile
environment that the outside world represents in an automotive application.
I would have thought that pulses of 120 v plus kicking about on the wiring
of a vehicle, would be the rule rather than the exception. Considering that
the cranking current of the starter motor can start out at over 100 amps,
and continue to be 40 or more until the engine starts, and that that load
then suddenly disappears as the starter is thrown out disengaging the
solenoid switch, is likely to cause significant spikes. Even a bit of a bad
plug lead, could cause large spikes to be induced in nearby wiring. I can't
actually imagine that any designer employed in the intensive world of
competitive automotive electronics, would not protect every rail and input
or output on his designs, with the cheap and effective devices available for
exactly this purpose, and if it is indeed the norm that this protection is
*not* designed in, then I am surprised that there are any cars still running
at all ...
Arfa